We are here to help. Contact us.

Adaptive Network Control and Management

In today’s computing and networking environment, system complexity has grown to the point where it can no longer be handled manually by human experts. There is a pressing need for automation which takes the system environment into consideration to determine appropriate actions. Researchers at Applied Communication Sciences are developing cutting edge technologies focusing on adaptive and cognitive systems in the following areas:

  • Policy-based network management: We have pioneered the field of policy-based network management for tactical wireless networks. As part of this work, we have spent the last decade developing advanced capabilities for automating network management based on policies. Our flagship policy-based network management tool, DRAMA, has been funded by the U.S. Army, Navy, and Air Force, demonstrating its applicability across all three branches of the military.
  • Fault diagnosis and remediation: We are developing capabilities for diagnosing fault and performance problems in tactical wireless networks, and for automatically reconfiguring the network to correct the diagnosed problems. The developed techniques make use of advanced machine learning algorithms for detecting and identifying anomalous events, and for learning the appropriate responses.
  • Automated reasoning: We are capabilities that allow automated analysis of policies by using automated reasoning technologies. These capabilities build upon Athena, a programming language and an interactive theorem proving environment rolled in one, which was developed by one of our researchers.
  • Dynamic spectrum access policy engine: We have created a unique policy engine for Dynamic spectrum Access (DSA) that combines expressivity and efficiency. DSA for cognitive radios is regulated by declarative policies specifying the conditions under which a transmission request is to be allowed. Policy engines for DSA need to reason about transmission requests in real time, as otherwise transmission opportunities may no longer exist, and may have to handle dozens of policies, each dealing with numerous parameters. Hence, efficiency is of paramount importance. Our policy engine has been shown to outperform all existing policy engines in this space